belief in the rationalizing logic of modernity as propounded by the development economists of the earlier generation, even though they differed significantly on whether the market or the central planner would be the agent of modernity. The subsequent rise of neo-Marxian scholarship restored class analysis to a central position. It was then assumed that class-based loyalties were in the end fundamental. In recent decades, however, social science disciplines have turned full circle. The class-centered approach has given way to new multi-dimensional accounts of identities that include ethnic, caste, religion, and gender categories, and thus social identities seem to have come to the centre-stage. In India, popular discourses show an overwhelming presence of issues around identities. However, on the changing salience of caste-based differences in the Indian context, there is a counter-position as well. Beteille (2012), for example, criticizes the ‘pre-occupation’ with caste that he observes among the experts who express opinions in the media on Indian affairs. He has argued that “rapid economic growth and the expansion of middle class are accompanied by new opportunities for individual mobility which further loosens the association between caste and occupation”. In other words, Beteille questions the material basis of the presumed persistence of caste-based differences in a rapidly changing Indian economy. Some of the recent studies have documented how the role of caste has been weaker than before in shaping economic well-being. Migration, expansion of lower and middle castes in non-traditional occupations, changes in agriculture, and most importantly affirmative action have all led to improvement in the relative position of the Scheduled Castes in India (Kapur et al., 2010). Yet, the political salience of the caste issue has hardly weakened. In the context of limited opportunities, overall and growing aspirations for upward mobility, competitively assertive caste identities for distributional gains continue to dominate the political discourse in India, the connection of which with the so-called material basis has been changing as a consequence.
Given the diversity of opinions regarding the importance of caste and other social groupings in Indian society, the question still remains what exactly has been happening to the distributional outcomes across social groups as a result of economic progress. Beteille (1983) made a useful distinction between two aspects of inequality — the relational and the distributional aspects. While sociologists and political scientists are mostly concerned with the first kind, economists are generally concerned with the second. In the first case, inequalities are seen as built into the social structure in the form of relations of super-ordination and sub-ordination, i.e. the patterns of privileges, rights, and obligations. An economist, on the other hand, sees inequality in the distribution of wealth or income, or, as Sen (1980) has proposed, in the distribution of human “functionings’ such as health or educational status. Why has the economist been rather less concerned about inequality across racial, ethnic, or caste groups? The answer probably lies in the methodological preference of the economist for a depersonalized agent as the unit of analysis. The agent acts independently to choose the best course of action given the opportunities. This way of thinking has definitely been fruitful in illuminating a variety of problems. It cannot, however, fully capture the ways inter-group inequality persists over time. There is no point in denying that one’s location within the network of social affiliations substantially affects one’s access to resources.
While the inter-relationship between economic development and economic inequality has long been explored by economists, and the earlier belief in an inverted U-shape between the two has been questioned in the light of extensive cross-country data for longer periods, there is very little analytical exploration into what might happen to inter-group inequality in course of rapid economic development. In this chapter, we first assess changes in measured inequalities between social groups in India in both income and non-income dimensions. In the process, we re-examine some of the issues in measurement of inter-group inequality, which would help us relook at inter-group disparities in other countries as well. Finally, we try to relate changing interpersonal and inter-group inequality to the fact that although some of the countries, such as Brazil, have not grown as fast as the Indian economy has grown, their record of reduction in inter-group inequality has surpassed India’s by a wide margin.
Why Between-Group Inequality
Relative disparities in well-being are often the concerns of the policymakers since sharpening disparities have the potential for creating conflicts in societies. Inequality between socially identifiable groups of people is considered ‘politically more salient and consequential than interpersonal comparison’ (Subramanian, 2011). Besides being considered intrinsically bad from an ethical standpoint, inequality between groups, what is often called ‘horizontal inequality’, has also been seen as having negative consequences on social coherence and peace (Stewart et al., 2005), even though the relationship between such inequality and the outbreak of violent conflict is not straightforward. There is perhaps an inverted U-shaped relationship between the two. When there is a large gap between the privileged and the disadvantaged groups, the disparity is either accepted with resignation or the privileged group might show some amount of magnanimity towards the less privileged so long as there is little threat from the bottom. The potential for conflict intensifies as the gap between the groups diminishes. While analyzing the occurrence of communal riots in India, Mitra and Ray (2014) find evidence in support of this hypothesis.
Between-group inequality also has an important bearing on the concept of inequality of opportunity. One way of assessing inequality of opportunity is the one that relies on the distinction between ‘circumstances’ and ‘effort’. While the unequal distribution of outcome across individuals may result from the differential efforts they put in to better their lives, it could also be due to the unequal circumstances they are in. The circumstances are defined to be those conditions which are beyond the control of the individual. In this conceptualization, individuals within a group are assumed to share the same circumstances and groups differ in terms of circumstances. If there was a systematic disparity between the average achievement levels of two different groups, it could be attributed to differences in circumstances, and as the argument goes, the focus of public policy should be on reducing the disparity due to circumstances.
In a society where individuals participating in the economic process are endowed with unequal quantities not only of economic assets but also of social assets, exclusion takes different forms. By social assets, we mean those ‘goods’ that belong to the realm of rights and entitlements. The fact that unequal access to publicly provided goods cannot be explained by the market process calls for such categories as social assets that include both political (e.g. citizenship rights) and cultural assets (i.e. ethnic markers such as ethnicity, religion, language, and so on). Access to employment, education, and productive assets, which is considered to be crucial in determining economic circumstances, varies across social groups in a manner which lies outside the control of an individual. Racial or caste identification of workers can interact with the social processes of human capital accumulation in communities and human capital valuation by employers in ways that generate externalities. In the presence of such externalities, as economic theory suggests, market processes may not lead to efficient outcomes. In the new economics of race, there is an emphasis on the links between under-investment in human capital due to circumstances and labor market outcomes.
Issues in Measurement
There has been a long stream of literature that has attempted to measure inequality across identifiable sub-groups using the method of ‘decomposition’. This approach views total inequality in personal income distribution as the sum of ‘between-group’ and ‘within-group’ components, and following this approach the between-group component is expressed as a percentage of the total inequality to reckon the contribution of between-group inequality to total inequality. Among the inequality measures, Theil’s entropy measure is commonly used for this purpose as it is readily decomposable, unlike the Gini coefficient — by far the most popular inequality measure — which is not decomposable in the way Theil’s measure is.
The general entropy class (Cowell and Jenkins, 1995) of measures is given by
